Introduction

Women’s relationship to and use of their bodies are issues of long-standing interest to feminist scholars. Despite the central role that exercise plays in many women’s lives—through its relationship with sports and athletics; with body image and (dis)satisfaction; and with mood and stress regulation—few researchers have addressed girls’ and women’s exercise behavior and the sociocultural context in which women make decisions about how, when, and why to engage in which types of exercise. For example, in the literature on body satisfaction, researchers have tended to focus on attempts to change one’s body via dieting rather than via exercise. Few researchers have asked about the lived experience of women or about the circumstances that shape their exercise-related decisions.

The studies reported here focus primarily on women’s participation in weight lifting, a form of exercise that is especially beneficial to women but often avoided by them. Because there is little prior work that addresses weight lifting, we approached the topic “from the bottom up”—by asking U.S. undergraduates at two mid-Atlantic colleges about their experiences in fitness centers—rather than by relying on a pre-existing theoretical framework. This enabled us to begin to formulate an account of women’s lived experiences. We gathered data in multiple ways: an archival analysis of Internet images (“Study 1”), closed-ended surveys (“Studies 2 and 3a”), and semi-structured essays (“Study 3b”). Before describing these studies and their goals in detail, we provide some background on gendered fitness regimens in the contemporary United States.

Gender-Typed Exercise and Fitness Ideals

Regardless of their gender, adults in the U.S. are enjoined to work out regularly. For example, consider Be Active Your Way, a guide produced by the President’s Council on Physical Fitness and Sports. It provides government-approved guidelines for physical activity: at least 75 min of vigorous or 150 min of moderate aerobic activity per week, plus muscle strengthening activity at least 2 days per week (Department of Health and Human Services 2008).

Despite the gender neutrality of the DHHS guidelines, the bodies to be produced by physical activity differ markedly for women and men in contemporary North America: masculine body ideals prescribe strength, while feminine body ideals prescribe thinness (Leit et al. 2002; Nichter 2000; Olivardia 2001). To match these ideals, men must build muscle (albeit just the right amount) while women must burn fat in order to become lean and toned (but not visibly muscular). These divergent gendered body ideals demand divergent exercise regimens. This article addresses one component of exercise regimens—weight lifting—that seems to be highly gendered.

Weight lifting is beneficial to the physical and mental health of both men and women. Well-toned muscles contribute to overall fitness and increase metabolism (Walberg 1989). They also help cushion the body from the effects of illness and slow the normal process of aging (cf. Angier 1999). Weight-bearing exercise is particularly important for women because after menopause, they lose bone mass more quickly than men. Weight-bearing exercise—particularly resistance training—can help prevent or forestall osteoporosis (Layne and Nelson 1999). Considered overall, lifting weights has clear biophysical benefits, some of which apply particularly to women.

There are also considerable psychological benefits of weight lifting for women. Research conducted on women in the United States suggests that weight lifting can provide a partial buffer against psychological conditions that are common among women. One example is eating difficulties (anorexia nervosa, bulimia nervosa, and subclinical manifestations of these disorders), which are far more prevalent among women and girls than men and boys. According to the Diagnostic and Statistical Manual of Mental Disorders, a key aspect of these difficulties is that body size and shape contribute disproportionately to self-evaluation (American Psychiatric Association 2000). Weight lifting may help to diminish body dissatisfaction. One study, for example, showed that 12 weeks of circuit training with weights improved women’s body satisfaction more than 12 weeks of aerobic training (Henry et al. 2006). Another study found that only 6 weeks of circuit weight training improved body satisfaction for both women and men (Williams and Cash 2001). Women are also at higher risk for depression than men, a pattern that emerges in mid-adolescence (Nolen-Hoeksema and Girgus 1994). Both aerobic and non-aerobic forms of exercise have been shown to improve mood; women in particular seemed to benefit (Rocheleau et al. 2004). Moreover, a short course of weight lifting improved the self-concepts of women who were clinically depressed and helped to alleviate clinical symptoms of depression as effectively as running (Doyne et al. 1987; Ossip-Klein et al. 1989).

Despite the benefits of weight lifting, everyday experience and observational studies in fitness centers and gyms suggest that women are usually underrepresented among users who lift free weights and engage in other types of weight-based strength training (Brace-Govan 2004; Khoury-Murphy and Murphy 1992). Why do women avoid weight lifting? This article affords a preliminary exploration of this question, focusing on young women in college settings in the United States. It focuses specifically on evaluation concerns that are intertwined with gender-related norms and ideals of femininity.

The term evaluation concerns refers to people’s interest in what others think of them. The need to belong, to be accepted, and not to be ostracized, is thought to be a primary human motivation, which is served by monitoring and managing social evaluations (Baumeister and Leary 1995). Evaluation concerns have been of perennial interest to social psychologists, who have concluded that such concerns are pervasive and powerful mechanisms of both self-regulation and social control. Feeling watched, scrutinized, or negatively judged by others can be strongly aversive. Social psychologists have typically studied how people cognitively and affectively “manage” this negative experience in laboratory settings. Outside of such controlled settings, however, people often respond to evaluation concerns through the simple expedient of avoiding settings and activities that might evoke such concerns. This may be what takes place in gyms and fitness centers: evaluation concerns discourage women from lifting weights there.

Our approach to evaluation concerns differs from the one often taken by social psychologists, who have typically focused on individual differences in the strength of evaluation concerns (Watson and Friend 1969; Leary 1983). Some have extended this focus to group-level differences. For example, Rudolph and Conley (2005) have argued that girls tend to experience stronger social-evaluation concerns than boys do, and that these concerns have both costs (such as girls’ greater likelihood of experiencing depression) and benefits (such as improved interpersonal competence). Both approaches differ from ours in construing concern about social evaluations as a relatively stable trait or feature of the individual.

In contrast, our approach examines how such concerns are evoked by specific situations or contexts. For example, evaluation concerns are readily evoked by situations in which a person engages publicly in behavior that is counter-normative. Counter-normative behavior is behavior that violates, or might be seen to violate, social expectations. Some norms apply to all members of a society (e.g., the norm to remain silent while watching a theater performance) but others apply selectively to members of specific groups. Gender is a social category that is strongly characterized by group-differentiated norms that specify how men and women should be and behave (see Prentice and Carranza 2002; Rudman and Fairchild 2004). Children appear to learn these gender-specific norms very young, through parental socialization (Raag and Rackliff 1998), and conform to them to gain approval (Banerjee and Lintern 2000). Sports participation is a domain in which normative expectations are very clearly gendered (Hardin and Greer 2009). However, to our knowledge there is no prior work on gender-specific normative expectations regarding gym-based physical exercise. We propose that such expectations exist; that women are aware of them; and that the expectations generate evaluative concerns for women that may contribute to their avoidance of certain types of exercise—and indeed, of gyms more generally.

The four studies reported below are intended to cast a wide net over these interlocked proposals. Study 1 examined the cultural background of associations between fitness goals and gender. Studies 2 and 3a examined specifically whether evaluation concerns are evoked by and feed back into lower rates of weight lifting in public. Study 3b examined women’s and men’s reports of critical incidents that resulted in discomfort in gyms. Together, this package of studies begins to shed light on the role of evaluation concerns in women’s pursuit of fitness.

Study 1

Are certain fitness goals linked to gender in popular imagery and discourse? Study 1 assessed the cultural associations that constitute the normative background against which women make exercise decisions. This aim complements past work on the sexualized or otherwise stereotypic ways that women are represented in popular visual media (e.g., Fink and Kensicki 2002). In a sense, the starting point was the reverse of past work: instead of seeking out representations of men and women and then analyzing the content of those representations, we sought out representations of certain fitness goals and then analyzed whether one gender was overrepresented in depictions of those goals. In other words, instead of asking “How is this gender represented?” the study asked “Who is depicted in the pursuit of this fitness goal?” Specifically, the study examined whether burning fat is a goal associated with women and building muscle is a goal associated with men. Furthermore, it examined trends in these associations over time.

To assess the presence and persistence of possible associations between gender and exercise goals, we carried out two Google Image searches each year for four consecutive years: one with the search term “burn fat” and the other with the search term “build muscle.” The resulting images are not a random sample from a population—rather, they are best understood as the most relevant images on the topic (as determined by Google’s criteria and algorithms). It would be inappropriate to infer or generalize from these images to some broader population. For this reason, we used descriptive rather than inferential statistics to analyze patterns in the data.

A researcher coded the gender of the person(s) depicted in the first 100 images returned by each search in March of 2007, and the first 200 images returned by each search in March of 2008, 2009, and 2010. For each search, the coder kept track of how many images depicted men only and how many images depicted women only. One coder was deemed sufficient because no subjective judgment was required. The remainder of the images depicted no people (e.g., depiction of a brand logo) or depicted both men and women. In addition, a small number of images were too indistinct to categorize.

The most direct index to answer the question “Who is depicted pursuing each fitness goal?” is a ratio. If burning fat is a goal associated with women, then we would expect to observe reliably more images depicting women than images depicting men. And if building muscle is a goal associated with men, then we would expect to observe reliably more images depicting men than images depicting women. In each case, the ratio of one gender to the other gender should be reliably greater than 1. Figure 1 illustrates these ratios annually from 2007 to 2010, and Table 1 provides the raw data on which the ratios are based. The key finding is that the ratios are greater than 1 for all years. In other words, both “burn fat” and “build muscle” are goals that are reliably associated with one gender.

Fig. 1
figure 1

Gender ratios depicted in Google images (“Study 1”).

Full size image
Table 1 Results of Google image searches (“Study 1”).
Full size table

These associations between gender and fitness goals constitute the cultural background—the shared social knowledge—linking women and men to different types of exercises. If the link between gender and type of exercise is strong and enduring, that link in and of itself may constitute a psychological barrier against participating in a type of exercise associated with the gender other than one’s own (at least in public). Thus, for example, women may avoid weight lifting in public because it seems to be an activity associated exclusively with men and therefore arouses concerns about negative social evaluation. A parallel idea has been raised in the popular press to explain men’s reluctance to take part in aerobics classes (Antrim 2005). In Study 2, we examined these ideas more closely by assessing women’s perceptions of exercise regimens.

Study 2

Study 2 examined whether perceived gender norms, in combination with evaluation concerns, constituted an obstacle to U.S. college women’s participation in weight lifting.

The first goal of Study 2 was to compare women’s perceptions of muscle-building exercise (lifting free weights using a bench press) to their perceptions of fat-burning exercise (using a StairMaster™). We assessed the women’s gender coding of these two activities. In line with the findings from Study 1, we predicted that using a bench press would be seen as characteristic of men, whereas using a StairMaster™ would be seen as characteristic of women (Hypothesis 1a). We also examined women’s exercise practices by asking how often they lifted weights and used the StairMaster™. We predicted that women would report lifting weights less often than they used the StairMaster™ (Hypothesis 1b). We further asked them to rate the extent to which each of these activities could help them to achieve their health and fitness goals. We predicted that women would report that their fitness goals were better served by fat-burning exercise than by muscle-building exercise (Hypothesis 1c). Asking women these questions built directly upon the previous study: whereas Study 1 assessed culture-level associations, Study 2 examined women’s perceptions of weight lifting and how those perceptions relate to their preferences and practices.

The second goal of Study 2 was to assess whether women anticipated that, were they to lift weights, they would risk receiving negative social evaluations. To examine whether lifting weights using a bench press would evoke evaluation concerns, we asked college women to imagine a scenario in which they were exercising either on a bench press or on a Stair Master™ and then to imagine a group of fellow students entering the fitness center. Half of the participants imagined male onlookers and half imagined female onlookers. We asked a series of questions to gauge participants’ anticipated reactions to others’ scrutiny. We predicted that women would report greater evaluation concerns when they imagined using a bench press than when they imagined using a StairMaster (Hypothesis 2). We made no specific predictions about whether onlookers’ gender would qualify this pattern.

Method

Participants and Design

Fifty-six female students from a private university on the east coast of the United States participated on the study. Most participants were passersby recruited in a campus dining hall and compensated $10 for completing this questionnaire among others. The remaining participants were recruited through the psychology department subject pool, which comprised all students enrolled in an early-level psychology course; these participants received partial credit toward a requirement for the course. In all cases, the study was described as being a questionnaire about the university’s fitness center. The final sample was about evenly divided between White (n = 29) and ethnic minority (n = 27) participants. Their mean age was 20.5.

Procedure and Measures

After giving consent, each participant was randomly assigned to imagine a public exercise scenario involving either weight lifting (n = 30) or aerobic exercise (n = 26). The weight lifting scenario read:

You are working out at the [university fitness center] (or another gym/fitness center that is familiar to you). It’s not long before closing, and there’s nobody else downstairs in the weight room. You want to bench press, but have no friends to spot you. You decide that you’ll just be careful not to lift more than you can handle safely. All of a sudden, several [guys/girls] come downstairs and start lifting a lot of weight on the bench press next to the one you were about to use. They seem really in shape.

The aerobic exercise scenario read:

You are working out at the [university fitness center] (or another gym/fitness center that is familiar to you). It’s not long before closing, and there’s nobody else on the cardio machines. You want to use the StairMaster, but wish you had some friends there so you wouldn’t be alone. All of a sudden, several [guys/girls] arrive and climb onto the StairMasters around the one you were about to use. They seem really in shape.

After reading the scenario, participants completed the measure of evaluation concerns. Next they answered questions about their perceptions of the exercises.

Evaluation concerns

Participants responded to five questions designed to measure evaluation concerns. Responses to the question “Would the fact that the other students had just arrived deter you from [lifting/exercising]?” were made on a scale bounded at 1 (definitely not) and 7 (definitely). Responses to the question “How comfortable would you feel relative to how you felt before they arrived?” were made on a scale bounded at 1 (much less comfortable) and 7 (much more comfortable). The last three items assessed concerns that the other students would be “watching you”; “evaluating you as a person”; and evaluating your performance on the task (i.e., “the amount of weight you were lifting” or “how good you were at the StairMaster”). Response options for the last three items ranged from 1 (totally unconcerned) to 7 (very concerned). Inter-item reliability for this composite index was good (Cronbach’s alpha = .89). We created an index of evaluation concerns by averaging the responses to five questions.

Perceptions of the exercises

Participants responded to four questions designed to assess the perceived typicality of men and of women doing the two exercises. They first rated how characteristic it was of women and of men to use a bench press. Then they rated how characteristic it was for women and for men to use a StairMaster™. Responses to the four questions were made on a scale bounded at 1 (not at all characteristic) and 7 (totally characteristic). Next, participants rated the extent to which they currently engaged in each of the two activities in order to achieve their health and fitness goals, with both ratings made on a scale bounded at 1 (not at all) and 7 (very much). Last, participants rated the extent to which each activity would be helpful in achieving their health and fitness goals, using scales bounded at 1 (not at all) and 7 (very much).

Results

Perceptions of the Exercises

As expected, none of the answers to the set of exercise perception questions was affected by the manipulations in the scenario portion of the study (all ps < .05).

Hypothesis 1a stated that using a bench press would be seen as a men’s exercise, while using a StairMaster™ would be seen as a women’s exercise. To test this hypothesis, we conducted a repeated-measures analysis of variance (ANOVA) on the typicality ratings with two within-subjects factors (target gender and activity). A crossover interaction strongly confirmed Hypothesis 1a, F(1,55) = 356.69, p < .001, η2 = .87. Participants rated using a bench press as much more characteristic of men (M = 6.20, SD = .75) than of women (M = 2.38, SD = 1.27), η2 = .89, whereas they rated using a StairMaster™ as much more characteristic of women (M = 5.79, SD = .91) than of men (M = 3.80, SD = 1.33), η2 = .66.

To test Hypothesis 1b, we examined participants’ reports of their actual use of the bench press and the StairMaster™. We used a paired-samples t-test to compare their ratings for the two types of exercise. Confirming hypothesis 1b, women reported using the StairMaster™ (M = 3.00, SD = 2.00) more than the bench press, (M = 1.77, SD = 1.51), t(55) = 3.99, p < .001.

To test Hypothesis 1c, we examined participants’ reports of how much the two exercises would be helpful in meeting their fitness goals. The hypothesis stated that women would say that aerobic exercise served their fitness goals better than weight lifting. Again, a paired-samples t-test was used to compare the ratings for the two types of exercise. Confirming Hypothesis 1c, women reported that the StairMaster™ (M = 4.84, SD = 1.69) would be more useful than weight lifting, (M = 2.89, SD = 1.81), t(55) = 6.05, p < .001.

Evaluation Concerns and Correlates of Evaluation Concerns

Hypothesis 2 stated that women would report stronger evaluation concerns when they imagined being observed using the bench press compared to being observed using the StairMaster™. We tested this hypothesis with an independent-samples t-test that examined the effect of type of exercise on the degree of evaluation concerns. (There were no differences due to the onlooker’s gender, so the two conditions were combined.) The analysis confirmed Hypothesis 2: participants reported greater evaluation concerns in the weight lifting condition (M = 4.56, SD = 1.58) than in the aerobic exercise condition (M = 3.75, SD = 1.27), t(53) = 2.08, p = .043.

If exercising, or even being in a gym, evoked strong evaluation concerns for women, then women might be motivated to avoid exercising. We cannot make a strong test of this prediction. However, we were able to examine whether the strength of evaluation concerns was related to shortfalls between women’s ideal level of exercise and their actual level. For this follow-up analysis, we calculated the difference between a woman’s estimate of ideal use and her estimate of actual use for each exercise. For the bench press, women with stronger evaluation concerns reported a greater disparity between their ratings of the usefulness of the exercise and their actual level of activity, r(55) = .29, p < .04. For the StairMaster™, there was no relationship, r(55) = .15, ns. This represents evidence, albeit indirect and modest, that evaluative concerns may uniquely discourage women from using the bench press.

Discussion

The results of Study 2 support the idea that evaluation concerns discouraged women from weight lifting. When participants imagined exercising in the presence of other people, those who imagined using a bench press reported greater evaluation concerns than those who imagined using the StairMaster™. Of course, we cannot say why this difference occurs. Two possibilities come to mind. One is that using a bench press more readily evokes evaluation concerns for women because it is a masculine-typed activity. Indeed, the “gender coding” of different types of exercise was unequivocal: participants saw the bench press as a men’s (and not women’s) exercise, while they saw the StairMaster™ as primarily a women’s exercise. Insofar as describing a gender difference can be tantamount to prescribing it (Rudman and Glick 2001), the masculine typing of the bench press may discourage women from weight lifting. A second possible reason why women reported low levels of use of the bench press involves evaluative concerns of a different sort. The college women we studied seldom used a bench press. Therefore, it is likely that they had not gained proficiency. In weight lifting, lack of proficiency is clearly visible to onlookers. Thus, a lack of proficiency may trigger evaluation concerns about looking inept. If so, one can envision how the two kinds of evaluation concerns combine in a self-perpetuating cycle: Evaluation concerns about gender-typing lead to infrequent use of the bench press, which in turn leads to low proficiency. Low proficiency produces additional evaluation concerns, which may lead to a further decrease in use. Such decreased use by women strengthens the gender-typing of the exercise. Studies that explore causal relationships directly are needed to test these speculations.

Study 2 was limited by its small sample size and the fact that we did not counterbalance the order of the questions. Additionally, one might question the comparison of the bench press and the StairMaster, two exercises that vary in a number of ways. By design, we chose to prioritize ecological validity instead of attempting to isolate one key variable that distinguishes these two activities. Rather than “purely” manipulating a single variable, this comparison accounts for the constellation of differences between the two primary forms of gym-based exercise available to women in real life. (See Gaertner et al. 1989 for a similar argument.) However, the specific elements most important to making weight lifting uncomfortable—the specific mechanisms behind the effects we observed here—should be addressed in future research. Future studies should also use larger sample sizes and should focus on actual behavior rather than imagined scenarios, as the next study does.

Study 3a

Study 3a was designed to address three questions about actual exercise experiences and behavior. First, do women report lower comfort levels than men in the gym? Second, do men and women report different patterns of gym equipment usage? Third, if so, do these gender-differences in usage patterns reflect the prevailing cultural associations we reported in Study 1?

To address these questions, we surveyed a group of male and female students at a small mid-Atlantic U.S. college about their use of different fitness equipment and their comfort in their college fitness center. The goal of Study 3a was to examine hypotheses about two forms of weight lifting: relative to men, women will report less frequent use of, and less comfort with, both weight machines (Hypothesis 1) and free weights (Hypothesis 2). As a point of reference, we also asked participants about their use of and comfort with aerobic equipment.

Method

Participants & Procedure

Participants were sixty-one undergraduates at a small liberal arts college on the east coast of the United States. The sample included 35 women and 26 men who were taking an introductory psychology course. Their mean age was 18.6 years. We did not gather data on ethnicity, but the college’s overall demographics are very similar to the sample from Study 2 (across the college, about half of students report their ethnicity as White).

All students in the course were invited to take part in a study for partial fulfillment of a course requirement, with a number of individual appointments available outside of class time. The invitation made no mention of fitness, exercise preference, or gender. The final sample comprised about 60% of those enrolled in the course. The remaining students were not available during the times the sessions were offered.

During the session, participants completed a packet of several questionnaires and scales administered by a female experimenter. The questionnaires pertinent to this study were among them and took less than 10 min to complete. Afterward, participants were debriefed with a short oral description of the study and its goals.

Questionnaire

The questionnaire contained several items assessing use of the college fitness center and its equipment, as well as an open-ended question about experiences of discomfort in the fitness center. Participants were first asked whether they had been to the college fitness center, and if so, how comfortable they felt there. Participants next reported how frequently they used three types of equipment (at any fitness center): “cardio” equipment for aerobic exercise (e.g., treadmills, bikes, and elliptical trainers); weight machines; and free weights (e.g., barbells or dumbbells). For each type of equipment, they circled one of the following options: never (coded as 1), once a year (coded as 2), once or twice a semester (coded as 3), once or twice a month (coded as 4), once or twice per week (coded as 5), and three or more times per week (coded as 6). Then those who circled any option except “never” were asked to rate their comfort level while using the equipment on a scale bounded at 1 (not at all comfortable) and 5 (very comfortable). A final open-ended item on the survey is discussed below. (See “Study 3b.”) The final part of the questionnaire asked participants to provide demographic information, including age, class year, and gender.

Results

Gym Use and Comfort Levels

We compared women’s and men’s reports of whether they had been to the college fitness center and, if so, how comfortable they felt there. Equivalent proportions of women (77%) and men (85%) reported having been to the college fitness center, χ2(1) = .53, ns. Of these 50 respondents, women (M = 3.37, SD = 1.31) rated themselves as less comfortable than men (M = 4.30, SD = .93), t(46) = 2.95, p < .01. Next, we compared men’s and women’s ratings of how frequently they used each type of exercise equipment and their ratings of how comfortable they felt using that equipment. (See Fig. 2 for a summary of use by women and men.)

Fig. 2
figure 2

Rates of fitness equipment use (“Study 3a”). Note. Women and men were equally likely to report having ever used aerobic equipment, X 2(1) = 0.27, ns. A greater percentage of men than women reported having ever used weight machines, X 2(1) = 3.57, p < 0.06, or free weights, X 2(1) = 4.11, p < 0.05.

Full size image

Weight machines

In support of Hypothesis 1, a gender difference emerged in the reported use of weight machines. Women’s level of use (M = 4.03, SD = 1.88) was lower than men’s (M = 5.00, SD = 1.20), t(56) = 2.43, p < .02. Of the 50 respondents who provided ratings, women (M = 3.38, SD = 1.27) rated themselves less comfortable than men (M = 4.33, SD = .92), t(48) = 3.01, p < .01.

Free weights

In support of Hypothesis 2, a gender difference emerged in the reported use of free weights. Women’s level of use (M = 3.03, SD = 2.04) was lower than men’s (M = 4.50, SD = 1.94), t(58) = 2.83, p < .01. Of the respondents who provided ratings (n = 38), women (M = 3.58, SD = 1.07) reported themselves less comfortable than men (M = 4.21, SD = .86), t(36) = 2.01, p = .05. Note that this comparison provides a conservative test of the gender difference in comfort. Non-users (of whom 75% were women) did not provide comfort ratings. It is likely that at least some of the nonusers avoided use because they anticipated discomfort.

Aerobic equipment

No gender differences emerged with respect to use of aerobic equipment. Women’s level of use (M = 4.24, SD = 1.78) was similar to men’s (M = 3.62, SD = 1.72), t(58) = 1.36, ns. Of the 49 respondents who provided ratings, women (M = 3.82, SD = .91) and men (M = 4.29, SD = .96) did not report significantly different levels of comfort while using the aerobic exercise equipment, t(47) = 1.74, p < .08.

Discussion

These data provide evidence that college women reported using weight training equipment less frequently than men. The self-report data are consistent with the commonplace observation that few women are found in the free-weight sections of fitness centers. The study also provided evidence that women reported being less comfortable than men in the gym. This was true in general, but particularly in their ratings of comfort when using weight machines and free weights. These self-report ratings provide direct evidence that discomfort is one aspect of many women’s experiences with gyms and fitness centers and with the use of weight machines and free weights in particular. However, we know little about the meanings that women themselves give to this discomfort, whether they attribute it to evaluation concerns, and how it compares to men’s (less frequent) discomfort in the gym. We now turn to Study 3b, in which the same participants were asked to describe experiences of discomfort in gym and fitness settings in their own words.

Study 3b

The final study analyzed critical incident reports written by participants in Study 3a in order to assess the sources of discomfort in gyms. The critical incident reports that we solicited concerned any uncomfortable incident that the participant had experienced in a gym setting—not just incidents involving weight lifting. (With relatively few female participants engaging in lifting weights, gathering a sufficient sample of incidents specific to weight lifting would not have been possible.) We solicited incidents through the final item on the survey used in Study 3a, which was an open-ended question worded as follows:

If you have ever felt uncomfortable or out of place at the [college fitness center] (or at your previous school’s gym), please describe the situation as completely as you can. Include what happened, who was involved, who else was there, how the situation was resolved, etc. Why exactly did you feel uncomfortable?

Note that this question was not addressed to individuals who lifted weights, nor did it ask specifically about lifting weights. Rather, as described above, the participants were a random group of students enrolled in an entry-level psychology class.

A total of 37 participants (61% of the sample) wrote about a critical incident in response to this item. All the incidents concerned emotional (not physical) discomfort. Although the question asked about a “situation,” many of the descriptions were not about a single discrete incident, but rather about ongoing or enduring difficulties. As one might expect, more women (25 out of 35, or 71%) than men (12 out of 26, or 46%) reported such a situation, χ2(1) = 3.99, p < .05.

We carried out a simple interpretive analysis of the critical incidents. The analysis was carried out by coders who remained blind (insofar as possible, given the content of the incident reports) to the gender of the writers. In order to capture the full range of participants’ responses to the open-ended question, we devised a two-step interpretive process. The first step was a “bottom-up” inductive analysis. That is, we sought to identify emergent themes that participants brought forward. Two coders (the first author and a male undergraduate, both white Americans), who were blind (insofar as possible given the content of the incident reports) to the gender of the writer, independently read the set of critical incidents. Using the method of constant comparison, each coder developed a set of categories (Glaser and Strauss 1967). These two coders then compared the two sets of categories. The comparison revealed close agreement, and the few small differences in wording were resolved through discussion. Three main categories emerged and the coders refined specific definitions of them, as follows:

  1. 1)

    concerns about evaluation by others (feeling scrutinized and/or judged by others);

  2. 2)

    concerns about comparison (judging oneself against the others, comparing oneself to others, and/or feeling that one does not measure up to others);

  3. 3)

    concerns about ineptitude (felt lack of experience or expertise).

The coders also noted a few ideas that arose only in one or two of the reports (such as feeling frightened in a “spooky” gym at night). These “orphans,” as Auerbach and Silverstein (2003) have termed them, were set aside.

In the second step, coders re-read the incident reports and, using the three categories noted above, identified the themes present in each incident report. The same two coders independently scored each incident report for the presence or absence of each of the three main themes. Thus, each incident report could be coded as making reference to one theme, more than one theme, or none of the themes. As during the first (category generation) step, the coders were not informed of the gender of the writers. Inter-coder reliability was assessed by examining percentage of agreement between the two original coders. Agreement was high for all three categories: evaluation concerns (86.5%), comparison concerns (86.5%), and ineptitude concerns (97.3%). The few disagreements between the two original coders were resolved by an independent additional coder (an English male graduate student).

Results

Gender was strongly related to the two of the themes found in the critical incident reports.

Evaluation Concerns

The theme of feeling watched, scrutinized, or judged by others was present in 12 of the 37 incident reports. Eleven of these incident reports were written by women (44% of the women who wrote an incident report) and only one by a man (8% of the men who wrote a report), χ2(1) = 4.71, p = .03.

Women’s evaluation concerns focused on both evaluations of bodily appearance (sometimes sexualized) and evaluations of competence. Regarding evaluations of bodily appearance, one woman wrote that in the free weight area of the gym, she found “exclusively men who stare at themselves in the mirror and try to act macho. I saw some of them look me up and down, which I hate.” Two women wrote about blatant evaluative comments. One wrote that men in the gym “talk loudly to each other, sometimes about the way that girls look.” Another woman wrote the following:

When it is about 4:00 and all the sports players arrive (I am not an athlete) it can be very uncomfortable to begin with. They talk loudly to each other, sometimes about the way that girls look. On one day, a friend of mine and I were getting a drink of water … and a guy talked about how “flabby” a girl’s legs were at the exercise bike. She didn’t seem so flabby. It’s hard to already feel uncomfortable, like it isn’t your place to use (i.e. only athletes can) and also know that people criticize the bodies of people in there when they are so vulnerable.

Some of the incidents that women described involved evaluation of their competence or performance. For example, one woman wrote, “I was afraid I looked inexperienced.” Another described her discomfort with using the chest press machine:

I would get all red and tremble trying to lift it. I was embarrassed just because other people in my class were around or waiting to use the machine, and I was struggling with the lowest weight.

Other women simply described evaluation concerns without specifying their nature. For example, one woman described walking into a fitness center where “there were people exercising. It was like they were all staring at me.” Another woman wrote, “There has never been a situation. I just am always scared that everyone is watching me.”

Comparison Concerns

The theme of upward social comparison to others (i.e., not measuring up) was mentioned in 16 of the 37 incident reports. Male participants accounted for proportionately more expressions of comparison concerns (66% of the men who wrote an incident report) than female participants (32% of the women who wrote an incident report), χ2(1) = 3.97, p = .05. For example, one man recalled comparing himself to the college’s star wrestler whenever they were in the fitness center at the same time: “I am uncomfortable because he is ridiculously big & can lift twice as much as me.” Another man wrote, “When much stronger, regular users of the weight room are present, such as football or lacrosse players, I get intimidated and feel pressure to be able to lift as much as them.” Two incident reports (both written by men) referred to overt competition, not just an internal process of comparison. For example, one wrote:

At my previous school we had physical education testing every semester and one test held in the weight room were pull-ups. The tests were done with one gender at a time and I always hated them because of the tacit or sometimes non-tacit competition between classmates to see who could crank out the most pull-ups. I never performed well and was never proud of that fact and so sometimes, I felt like my manhood was a mockery.

A woman reported an incident in which she retreated to the outdoor track after entering the fitness center and realizing that “the people inside were all in shape and I wasn’t.”

Ineptitude Concerns

A theme of lacking experience was present in a total of 8 of the 37 reported incidents. There was no gender difference in the mention of this theme: 33% of the men who wrote a report mentioned it, as did 16% of women, χ2(1) = 2.22, ns. Examples of ineptitude concerns include the following: “I did not really know what I was doing, and was slightly intimidated by other people lifting 5× more than I was.”

Discussion

Women responded more frequently than men to the invitation to write about a situation in a gym that made them uncomfortable. Moreover, women were more likely than men to write about incidents involving feeling or being scrutinized. Their concerns focused on social evaluations of both body weight/shape (sometimes sexualized) and lack of competence and skill. Only one man recounted an incident involving concern about social evaluation. The gender difference suggests that evaluation concerns influence women’s experiences in the gym.

General Discussion

Ideally both men and women would be comfortable being in gyms and using all of the equipment in them. But this series of studies showed that many college women’s experiences with weight-based exercise (and in the gym more generally) were marked by psychological discomfort, heightened concern about social evaluation, and avoidance of certain activities.

As we noted in the introduction, engaging in load-bearing, weight-based exercise has important health benefits for women. However, relatively few of the college women we studied reported lifting weights, an activity that many regarded as typical of men. Why did women avoid lifting weights and other weight-based strength training? The studies reported here, along with popular accounts and previous scholarship, offer some tentative answers. One answer is that the cultural ideal of feminine bodies discourages women from weight lifting. Angier (1999) reported that she frequently heard women say that they “wouldn’t want to bulk up or to look too muscular” (p. 293). Similarly, Khoury-Murphy and Murphy (1992), who conducted an ethnographic study of a program of weight lifting for older women, reported that these women cited gaining too much muscle as an undesirable outcome of weight lifting.

A second factor that may contribute to women’s reluctance to exercise with weights in public is the perception that weight lifting, along with some other types of exercise, is widely regarded as a masculine activity. Among the college-age women we studied, weight lifting was seen as typical of men, not women. Other researchers have shown that gyms (or particular areas of gyms) are male preserves (cf. Brace-Govan 2004). Yet other researchers have shown that certain sports (e.g., rugby; Fallon and Jome 2007) are seen as men’s sports (see also Hardin and Greer 2009).

The studies reported here identified another barrier that might keep women from engaging in weight lifting: evaluation concerns. There are several interrelated ways that using free weights and a bench press might evoke women’s evaluation concerns. As the data showed, using a bench press was perceived as a masculine activity; moreover, women were relatively unlikely to exercise with free weights. Engaging in these forms of exercise would violate gender expectations (see also Salvatore 2002, Study 2). Furthermore, the proper use of weights and a bench press requires considerable skill; a novice user would be more likely to use the equipment improperly, thus triggering evaluation concerns about competence. Bench press proficiency—both skill (form) and strength (amount of weight lifted)—can be readily assessed by onlookers. These issues of apparent competence pertain to all weight lifters, both men and women. However, as “Study 3a” showed, women were more likely than men to be novice or infrequent users; thus, women may be more likely both to lack the skill and strength that comes with practice. As a result, their concerns about displaying low competence would be stronger.

Taking these factors together, we conjecture that women’s relationship with weight lifting may involve a number of self-perpetuating cycles. First, women may avoid gyms (or weight training areas of gyms) because they see them as masculine spaces; the absence of women reaffirms the gender coding of those spaces and thus, serves to perpetuate women’s avoidance of them. Similarly, many women may avoid weight lifting because they see it as a masculine activity; the small number of women lifting weights in turn reaffirms the gender coding of that activity. Furthermore, if women seldom lift weights, they will not become proficient with the equipment; lack of proficiency contributes to concerns about being seen as incompetent or inept. These conjectures are only speculative at this point. They need to be tested with further research that uses real-life manipulations (rather than imagined scenarios and paper-and-pencil measures) and direct assessments of the possible causal mechanisms, and outside of college convenience samples.

It is premature to propose specific ways of encouraging women to include weight lifting in their fitness regimens. However, the data suggest that certain strategies will be more successful than others. Providing accurate information about the health benefits of weight lifting is likely not sufficient, because women’s decisions not to lift weights are not based solely on an instrumental cost-benefit analysis. It is important to address their evaluation concerns. One might suggest that opportunities by created for women to lift weights in private. However, such a solution would likely be available to only a limited number of women—those who could afford costly equipment and have reliable access to private space. More important, private exercise is a private solution; it leaves cultural gender ideologies intact. Alternatively, one might suggest psycho-educational technologies to alter “faulty” cognitive schemata that lead women to be overly concerned about social evaluation. This strategy, however, implicitly holds women responsible for problems that are social and cultural in nature (see also Crawford and Marecek 1989, on the “woman as problem” in psychological theorizing).

The critical incidents that women reported suggest that their evaluation concerns often were produced and sustained by everyday social relations in gyms, including objectifying discourses, overt evaluative commentary by men, and exclusionary practices. Therefore, we favor interventions that aim to interrupt such practices. For example, women who are relatively experienced lifters might work to build a critical mass of women lifters, by recruiting novices and helping them gain skills. If groups of women lifters become increasingly visible in a gym, this may counter the implied proscription against women lifting. Moreover, groups of women acting collectively might feel empowered to “call” men on sexist commentary or other forms of exclusionary treatment or to ask for formal institutional redress of such behavior.

Since the passage of Title IX in 1972, there has been no formal limitation on female students’ access to the full array of exercise and fitness options found on college campuses. As a result, many more women participate in college athletics and make use of athletic facilities than before. Yet, informal barriers to women’s full participation remain. College gyms remain gendered spaces and some types of exercises and some sports remain highly gender-typed. As in the case of weight lifting, these barriers may deprive women of important health benefits. This underscores the need to understand the cultural norms, everyday informal practices, and gendered social relations that constitute the social world of the gym.